IT/Software career thread: Invert binary trees for dollars.

Tuco

I got Tuco'd!
<Gold Donor>
46,831
78,477
Why matlab for prototyping? Its syntax would make me want to skip prototyping altogether and go straight to C/C++.
It's pretty hard to beat matlab in terms of number of lines and ease of use when doing linear algebra, which makes up a great deal of the mathematics in robotics. But I think it's mostly widely used because academics like it.

Personally I use C++ for nearly everything just because I want even my prototypes to run in real time.
 

Cad

scientia potentia est
<Bronze Donator>
25,344
48,546
lol I felt the same way, OOP has no principals, OOP "principals" are re-defined on the fly as it suits the convenience of the person arguing them.
rrr_img_119910.jpg


Principal vs. principle - Grammarist
 

Tuco

I got Tuco'd!
<Gold Donor>
46,831
78,477
Cad, what is the most principle principal in OOP?
edit: let it be known I made my principle/principal joke before I saw Cad's post.
 

Asshat wormie

2023 Asshat Award Winner
<Gold Donor>
16,820
30,968
It's pretty hard to beat matlab in terms of number of lines and ease of use when doing linear algebra,
Numpy/Scipy > Matlab
smile.png


Also no way was a_skeleton_03's class an actual algorithm/data structure class. Looks like CS 101.
 

Lendarios

Trump's Staff
<Gold Donor>
19,360
-17,424
Siuneg, I hoped you liked the class and learned how to separate a complex problem into small solvable items. That is the principal of programming in a nutshell.
 

Noodleface

A Mod Real Quick
38,245
15,030
Spent all day yesterday and continuing to merge two BIOS code bases (based off a single codebase, but with their own modifications). I've never written so many #ifdefs in my life.

Still a few hundred files to modify by hand. Hoping once it's done the BIOS images still work
smile.png
 

TJT

Mr. Poopybutthole
<Gold Donor>
42,516
108,165
Just finished an application for my team I had to rush out the door in like 1.5 months. I develop financial software for internal use at a large corporation. Most of it has to do with ETL shit and database design.

Our team had been having difficulty conveying to leadership some of our work due to using multiple project management systems and some other issues so despite being extremely effective in what we do. We weren't getting this across as well as we'd like. As we don't game the numbers, like some other teams. Over the summer I had our intern build a suite of really awesome dynamic reports for this purpose. Only to have our Tool Distribution team deny us the permissions to use them. Despite being a native functionality of our primary management system (lolredtape).

Because of all that, I had to make a workaround application to make this reporting work. Got an unused desktop box from the IT dudes. Hooked it up and turned it into our own VM/Database. Designed a DB that reorganized all the data more efficiently. Then wrote an application to pull all of our project manaement metrics from three different management platforms into the DB I created.

Modified the intern's report suite to work with the new local DB and then created an Access Application for the rest of the team to input their comments for project based reports and such. As the original Sharepoint concept I had, I kept having to refactor like every week and half due to the constant undocumented changes that team makes to platform. Gag.

Fun little project though.
 

moontayle

Golden Squire
4,302
165
A fix for my timing issue was merged into the main branch of the library so now I just need to wait for that to make its way into the world at large.

Project is still in testing. No feedback so far. Doing other things in the meantime.
 

ShakyJake

<Donor>
7,885
19,828
So, back in 2008 our company re-developed our bread-and-butter application from an old school VB6 desktop app to Silverlight. Big mistake because we're now paying for it as more browsers disallow the plugin. We now have a team investigating which technologies and frameworks to rebuild the application with.

The team responsible is clearly heading us towards a Javascript client-side framework such as Angular, React, or Ember and a WebAPI back-end probably using ASP.NET WebAPI2 and MSSQL (which makes sense because we're not redoing the db).

I have my reservations on the Javascript front-end, though. It makes sense for traditional public-facing web applications, but for us our application is used within hospital systems. I've been warning them that it may not be a good idea to have our front-end written entirely in Javascript. It's possible a hospital's IT could disable Javascript support on their browsers for security reasons. If they were to do this then our application wouldn't work. Period.

Additionally, our developers have little to no Javascript experience. Trying to apply C# and OO constructs to Javascript will be a mess.

I'm suggesting they go the ASP.NET MVC route and just spit out plain-jane HTML and CSS withminimalJavascript (for validation and some nice UI-widgets like date-pickers, etc.). So Javascript isn'trequiredand the application will continue to function if it's turned off.

I get why they want to completely separate the front-end from the back. By going ASP.NET MVC route we could be facing the same dilemma 5 years from now, I suppose.

I dunno, just kinda wanted to rant a bit.
 

moontayle

Golden Squire
4,302
165
My thought would be to err on the side of caution like you're suggesting. I mean, the decision makers could always negotiate not having clients turn off JS but that's a headache for both sides and the client could always go elsewhere if they felt security was at risk. Shit, they should KNOW this already. I've been in the hospital industry. HIPAA dictates a lot, including security. I guarantee that's a non starter for a lot of places.
 

ShakyJake

<Donor>
7,885
19,828
I've got a whopper of a project we're working on for this release and I seriously don't believe it's feasible.

First, some background: as I mentioned previously, our product is used in hospital systems. One of the features of this product is that we can transmit HL7 data to the hospital's HIS. Up to this point the mechanism that we used to massage data into a form that the HIS accepts is via an old-ass Visual Basic 6 application. We have an entire department whose function is to modify VB6 scripts to transform the data into whatever the client needs.

It's works great and we can pretty much do whatever the client requires since we have the full power of a programming language (e.g. conditionals, loops, wealth of string formatting options, etc.) Problem is, it being VB6 is starting to cause problems and it's really at a point where it needs replacing. What sucks is that any replacement will require touching every deployed system out there since the code is specific to Visual Basic 6. There's no porting it to straight into a .NET version or anything. Basically we're looking at a complete rewrite for every site. It sucks and there isn't any way around that.

However, for it's replacement TPTB wants a dumbed-down GUI-based tool that will allow formostof the functionality of the VB6 app. Conceptually, the tool will have drop-down pick-lists for "IF THEN ELSE", "STRING CONCAT", "STRING APPEND" -- shit like that. In my opinion, this is insane. Absolutely insane. MAYBE this would be feasible if we had real UI experts working for us, but we don't. Most of our devs are back-end guys with minimal UI experience. In fact the demo tool that the team spent months working on is pretty ugly. It actually works okay for simple scenarios but it quickly explodes into an unwieldy mess once one needs to do nested conditionals and stuff (which is a pretty common requirement).

Anyway, I'm wondering if anyone has had experience with any sort of tool that does this sort of thing (i.e. a UI-based script engine). I really would like to see some examples of their approach. I also toyed with the idea of creating our own VERY simple scripting language but, man, I have no idea where to begin on that one. And I realize that itself is no simple task.

Lastly, don't even attempt to recommend third-party solutions such as Mirth. Not-Built-Here-Syndrome is strong with us.
 

Noodleface

A Mod Real Quick
38,245
15,030
Some guy at work made a gui scripter for ssh sessions here since everything is done over serial comm. He did it in a way that you'd make a block diagram of your scripting steps and the program converted it to iron python and ran the script. Whole company adopted it.

The program SUCKED DICK. I hated it and never used it.
 

ShakyJake

<Donor>
7,885
19,828
Some guy at work made a gui scripter for ssh sessions here since everything is done over serial comm. He did it in a way that you'd make a block diagram of your scripting steps and the program converted it to iron python and ran the script. Whole company adopted it.

The program SUCKED DICK. I hated it and never used it.
Yeah, we're going to face a mutiny when the implementation guys see our proposal. Again, they're used to the flexibility of a programming language in order to get the job done. Forcing them to use a UI-based scripting "language" is not going to go over well.

What I suggested in the past (and was promptly dismissed out-of-hand) was to use an embedded scripting language, such as Javascript, Python, or even Lua, that the engine would execute against some context object. This way, the scripts themselves have no dependency on the framework they're executed on. We could easily switch from .NET to Java or whatever and the scripts will still be valid.